Optimized sampling design and rationale for verification and validation

Abstract

An optimized sampling design that meets customer, design, or process requirements, while balancing technology limitations, is still a common challenge to engineering communities. This is especially true in the medical device industry. Acceptance sampling plans for manufacturing are widely available, but the appropriate sampling plans for verification and validation (V&V) are less well known. This paper applies established statistical theory to derive sampling plans appropriate for estimating product reliability during V&V, where reliability must exceed an established threshold with an appropriate margin of statistical confidence. The paper provides insight on how to estimate parameters of interest and interpret acceptance criteria. Operating characteristic curves are used to examine if a design or process is capable of producing future product that meets design specifications and/or customer requirements in terms of confidence and reliability. The methodology is applied to both attribute and variable sampling plans, including examples showing how to achieve a high probability of passing the acceptance criteria. Formulas, sample size tables, and operating characteristic curves are provided for engineering practitioners to use. The paper aims at providing a practical quantitative approach and a valid statistical rationale to assess overall product quality during V&V.